ICML 2022spotlight29 citations
Shuffle Private Linear Contextual Bandits
Sayak Ray Chowdhury, Xingyu Zhou
Abstract
Differential privacy (DP) has been recently introduced to linear contextual bandits to formally address the privacy concerns in its associated personalized services to participating users (e.g., recommendations). Prior work largely focus on two trust models of DP – the central model, where a central server is responsible for protecting users’ sensitive data, and the (stronger) local model, where information needs to be protected directly on users’ side. However, there remains a fundamental gap in the utility achieved by learning algorithms under these two privacy models, e.g., if all users are
BibTeX
@InProceedings{pmlr-v162-chowdhury22a,
title = {Shuffle Private Linear Contextual Bandits},
author = {Chowdhury, Sayak Ray and Zhou, Xingyu},
booktitle = {Proceedings of the 39th International Conference on Machine Learning},
pages = {3984--4009},
year = {2022},
editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan},
volume = {162},
series = {Proceedings of Machine Learning Research},
month = {17--23 Jul},
publisher = {PMLR},
pdf = {https://proceedings.mlr.press/v162/chowdhury22a/chowdhury22a.pdf},
url = {https://proceedings.mlr.press/v162/chowdhury22a.html},
abstract = {Differential privacy (DP) has been recently introduced to linear contextual bandits to formally address the privacy concerns in its associated personalized services to participating users (e.g., recommendations). Prior work largely focus on two trust models of DP – the central model, where a central server is responsible for protecting users’ sensitive data, and the (stronger) local model, where information needs to be protected directly on users’ side. However, there remains a fundamental gap in the utility achieved by learning algorithms under these two privacy models, e.g., if all users are